Forest Monitoring of the Congo Basin using Synthetic Aperture Radar (SAR)

1
Forest Monitoring of the Congo Basin using
Synthetic Aperture Radar (SAR)

• James Wheeler
• PhD Student
• Supervisors Dr. Kevin Tansey, Prof. Heiko
  Balzter
• Department of Geography
• University of Leicester
• james.wheeler_at_le.ac.uk

2
Outline

• Congo Basin Background
• Congo Rainforest and REDD
• Remote Sensing for monitoring deforestation
• Optical
• Radar
• Objectives

3
Congo Basin Background

• Area drained by the Congo River
• 3.8 million km2
• 1.5 million km2 tropical forest
• Swamp forests mountain ecosystems
• 2nd largest rainforest

4
Congo Rainforest and REDD

• Reducing Emissions from Deforestation and Forest
  Degradation (REDD)- requires Measurement,
  Reporting and Verification (MRV) system.
• Congo contains approx. 46 billion Mg Carbon
• Rate of forest loss in Central Africa 0.2/year
  (Duveiller et al. 2008)
• Change occurring on fine scale spatial
  resolution for monitoring ideally 100m or higher

5
Measuring Deforestation in the Congo

• Ground based measurements only 300 Ha of
  permanent sample plots in Congo Basin (Nasi et
  al. 2010)
• Useful for validation of remote sensing methods
• Forest inventory data important for modelling
• Remote sensing methods
• Optical
• LIDAR
• Radar

6
Rainforest Biomass Data Uncertainty
Biomass study using satellite borne LIDAR dataset
by Saatchi et al. (2011)
7
Optical Remote Sensing

• Measure reflectance of visible and near infrared
  sunlight
• E.g. Landsat datasets
• Freely available
• Continuous time series for gt30 years
• Affected by cloud/haze
• Can be useful for decadal/semi-decadal
  assessments
• Easier to visually interpret

8
Optical Remote Sensing of the DRC

• Image from Hansen et al. (2008)

9
Radar Remote Sensing

• Synthetic Aperture Radar (SAR)
• Penetrates Cloud, not affected by atmosphere/haze
• Active system can operate at night
• Measures geometric and dielectric properties
• Surface roughness
• Moisture content

10
Image from Canada Centre for Remote Sensing
11
Radar Remote Sensing

• Common radar band wavelengths
• X 3cm C 5.6cm L 23cm P75cm
• X-band data available from Tandem-X mission
• C-band data available from Envisat, Radarsat,
  future Sentinel-1
• L-band data available from 1992-1998, 2006-2011
• Currently no P-band satellite data
• Possible future BIOMASS mission

12
Radar

• Longer wavelengths (L- and P-) appropriate for
  forestry applications higher penetration into
  canopy
• Multiple modes of acquisition benefit forest
  monitoring
• Polarisation of emitted and received signal can
  be used to measure different characteristics of
  target
• Unfortunately no satellite L-band data from April
  2011 until at least 2013

13
Radar Reflectance
Image from Canada Centre for Remote Sensing
14
Kyoto and Carbon Initiative PALSAR Mosaic

• L-band
• Long data strips
• Data collected over 2 months

• De Grandi et al. 2011

15
Multi-sensor approach

• E.g. Congo wetlands mapping project
• Optical data
• L-band radar
• Digital Elevation Model

Image from Bwangoy et al. (2010)
16
Current Objectives

• Using L-band SAR data
• Produce map of forest loss from analysis of
  backscatter, signal coherence
• Combine different techniques in a multi-sensor
  approach
• Field expedition to Republic of Congo 2012/2013
• Develop a method for biomass estimation from
  satellite data to complement knowledge about
  deforestation

17
References

• Bwangoy, J.-R.B. et al., 2010. Wetland mapping in
  the Congo Basin using optical and radar remotely
  sensed data and derived topographical indices.
  Remote Sensing of Environment, 114(1), pp.73-86.
• Canada Centre for Remote Sensing.
  http//ccrs.nrcan.gc.ca/resource/tutor/gsarcd/inde
  x_e.php
• Duveiller, G. et al., 2008. Deforestation in
  Central Africa Estimates at regional, national
  and landscape levels by advanced processing of
  systematically-distributed Landsat extracts.
  Remote Sensing of Environment, 112(5),
  pp.1969-1981.
• De Grandi, G. et al., 2011. The KC PALSAR mosaic
  of the African continent processing issues and
  first thematic results. IEEE Transactions on
  Geoscience and Remote Sensing, 49(10),
  pp.3593-3610.
• Hansen, M.C. et al., 2008. A method for
  integrating MODIS and Landsat data for systematic
  monitoring of forest cover and change in the
  Congo Basin. Remote Sensing of Environment,
  112(5), pp.2495-2513.
• Nasi, R. et al., 2010. Carbon Stocks and Land
  Cover Change Estimates in Central Africa - Where
  Do We Stand? In M. Brady C. de Wasseige, eds.
  Monitoring Forest Carbon Stocks and Fluxes in the
  Congo Basin. Brazzaville, Republic of Congo, pp.
  10-13.
• Saatchi, S.S. et al., 2011. Benchmark map of
  forest carbon stocks in tropical regions across
  three continents. Proceedings of the National
  Academy of Sciences of the United States of
  America, 108(24), pp.9899-904.

18
Thank you for listening
19
(No Transcript)